--- title: Why do customers choose Langfuse? description: Learn why we built Langfuse and what makes it different from other AI engineering platforms. --- # Why do customers choose Langfuse? Customers choose Langfuse because we are... - The most used open-source AI Engineering platform [(blog post)](/blog/2024-11-most-used-oss-llmops) - Model and framework agnostic with [100+ integrations](/integrations) - Built for production & scale - Designed for complex agents and multi-step workflows - Offering a complete toolbox for AI Engineering - Incrementally adoptable, start with one feature and expand to the full platform over time - API-first, all features are available via API for custom integrations - Based on [OpenTelemetry](/integrations/native/opentelemetry) for interoperability - Easy to [self-host](/self-hosting) Langfuse is the most widely adopted LLM Engineering platform: - Used by **50,000+** companies - **33,972** GitHub stars - **130M+** SDK installs per month - **6M+** Docker pulls - Trusted by **21 of the Fortune 50** and **129 of the Fortune 500** Companies who trust Langfuse: ## Open Source - Langfuse is [open source](/handbook/chapters/open-source). - You can [self-host](/self-hosting) it, including the scalable observability backend that powers Langfuse Cloud without scalability limitations. - All product capabilities—tracing, evaluations, prompt management, experiments, annotation, the playground, and more—are MIT licensed without any usage limits. - We are transparent about [what's on the roadmap](/docs/roadmap) and [how we operate](/handbook). - We iterate with the [community on GitHub](https://github.com/orgs/langfuse/discussions) because best practices for building great LLM applications are rapidly evolving. Please share your feedback and ideas with us. - Self-hosting keeps data within your infrastructure; Cloud offloads operational overhead. You can switch between OSS, Enterprise self-host, and Langfuse Cloud at any time—no feature flags to untangle, no vendor lock-in. ## Extensive Integrations built on OpenTelemetry - Native SDKs for [Python](/docs/observability/sdk/python/overview) and [JavaScript/TypeScript](/docs/observability/sdk/typescript/overview) - [100+ integrations](/integrations) with popular frameworks, model providers, and tools - Framework support: [LangChain](/integrations/frameworks/langchain), [LlamaIndex](/integrations/frameworks/llamaindex), [OpenAI Agents](/integrations/frameworks/openai-agents), [Vercel AI SDK](/integrations/frameworks/vercel-ai-sdk), [CrewAI](/integrations/frameworks/crewai), and many more - Model providers: [OpenAI](/integrations/model-providers/openai-py), [Anthropic](/integrations/model-providers/anthropic), [Google Gemini](/integrations/model-providers/google-gemini), [Amazon Bedrock](/integrations/model-providers/amazon-bedrock), and others - LLM Gateways: [LiteLLM](/integrations/gateways/litellm), [OpenRouter](/integrations/gateways/openrouter), [Portkey](/integrations/gateways/portkey) - Native [OpenTelemetry](/integrations/native/opentelemetry) support for maximum interoperability ## Developer First - Langfuse is built for developers. - We are creating a technical product with great developer experience. - Langfuse is powerful yet simple, allowing you to build custom logic on top of it. - All data is accessible via [public APIs](/docs/api-and-data-platform/features/public-api) and [SDKs](/docs/api-and-data-platform/features/query-via-sdk). - [MCP Server](/docs/api-and-data-platform/features/mcp-server) for AI-native workflows and integrations. ## Reliable Partner - All changes to the Langfuse API and integrations are covered by [semantic versioning](https://semver.org); we test public interfaces to ensure backward compatibility and run end-to-end integration tests for the most common use cases in CI. - Raised a [$4M seed round](/blog/announcing-our-seed-round) from Lightspeed Ventures, General Catalyst, Y Combinator, and angel investors. - [Langfuse has been included twice in the Thoughtworks Tech Radar](https://www.thoughtworks.com/en-us/radar/platforms/langfuse) as a recommended platform. - Strong adoption and community growth ([see metrics](#public-metrics)). - [Learn more](/about) about us as a team. ## Built for Complex Use Cases - We designed Langfuse with complex, nested LLM calls in mind—especially for agents and multi-step workflows. - [Agent graphs](/docs/observability/features/agent-graphs) provide visual representations of complex agent workflows, helping you understand and debug multi-step reasoning processes. - Langfuse enables hierarchical representations of your application in [traces](/docs/observability/overview). Why are traces the core abstraction for LLM Engineering? Learn more in this [webinar](/resources/engineering/webinar-observability-llm-systems). - [Multi-modal support](/docs/observability/features/multi-modality) for tracing text, images, audio, and other modalities. - [MCP tracing](/docs/observability/features/mcp-tracing) for Model Context Protocol server interactions. - We go beyond Input/Output to include all the context of your app via [metadata](/docs/observability/features/metadata), [tags](/docs/observability/features/tags), and [sessions](/docs/observability/features/sessions). ## Comprehensive Platform We are building the core development platform you need to build robust LLM applications. Langfuse offers four integrated pillars: - **[Observability](/docs/observability/overview)**: Comprehensive tracing for LLM applications, including [agent graphs](/docs/observability/features/agent-graphs), [sessions](/docs/observability/features/sessions), [token & cost tracking](/docs/observability/features/token-and-cost-tracking), and [multi-modality](/docs/observability/features/multi-modality). - **[Prompt Management](/docs/prompt-management/overview)**: Version control, [LLM Playground](/docs/prompt-management/features/playground), [A/B testing](/docs/prompt-management/features/a-b-testing), [GitHub integration](/docs/prompt-management/features/github-integration), and collaborative workflows. - **[Evaluation](/docs/evaluation/overview)**: [LLM-as-a-Judge](/docs/evaluation/evaluation-methods/llm-as-a-judge), [human annotations](/docs/evaluation/evaluation-methods/annotation), [experiments](/docs/evaluation/core-concepts#experiments), and [datasets](/docs/evaluation/experiments/datasets) for systematic testing. - **[Metrics & Data Platform](/docs/metrics/overview)**: [Custom dashboards](/docs/metrics/features/custom-dashboards), [metrics API](/docs/metrics/features/metrics-api), and extensive [data export options](/docs/api-and-data-platform/overview). ## Built for Scale ## Security and Compliance - We take security and compliance seriously. - **Certifications**: Langfuse Cloud is [SOC 2 Type II](/security/soc2) and [ISO 27001](/security/iso27001) certified. - **Privacy**: [GDPR](/security/gdpr) compliant with [DPA](/dpa) available. [HIPAA](/security/hipaa) aligned via a dedicated cloud region and signed Business Associate Agreement (BAA) for eligible customers. - **Data Regions**: Choose between US, EU, or HIPAA-ready data regions on Langfuse Cloud—or [self-host](/self-hosting) anywhere. - **Data Control**: [Data masking](/docs/observability/features/masking), [data retention](/docs/administration/data-retention), and [data deletion](/docs/administration/data-deletion) capabilities. - More details in our [Security & Privacy Center](/security). ## Public Metrics [View the Langfuse public metrics dashboard](https://app.hex.tech/clickhouse-analytics/app/0349rZiOtG8QDXATQrRCSI/latest). --- ## Agent Instructions This page is part of the [Langfuse](https://langfuse.com) documentation, published as plain Markdown for AI agents. Every page is available as Markdown by appending `.md` to its URL, or by sending an `Accept: text/markdown` header. This page: `https://langfuse.com/handbook/chapters/why.md`. ### Querying these docs If the answer is not on this page, query the documentation instead of guessing: - **Semantic search** across all Langfuse docs, returning an answer with the relevant pages and excerpts. Ask a specific, self-contained question: ```bash curl -sG "https://langfuse.com/api/search-docs" --data-urlencode "query=How do I trace a LangGraph agent?" ``` - **Index of every page**: , with per-section indexes [llms-docs.txt](https://langfuse.com/llms-docs.txt), [llms-integrations.txt](https://langfuse.com/llms-integrations.txt), and [llms-self-hosting.txt](https://langfuse.com/llms-self-hosting.txt). ### Before writing Langfuse code - **Install the [Langfuse Agent Skill](https://langfuse.com/docs/api-and-data-platform/features/agent-skill).** It encodes Langfuse's own best practices for instrumentation, prompt management, and evaluation, and materially improves results. - **Read [What does a good trace look like?](https://langfuse.com/docs/observability/best-practices.md)** before instrumenting an application. - **Verify endpoints, parameters, and response fields** against the [API reference](https://api.reference.langfuse.com) instead of inferring them from code examples. - **Use the [Langfuse CLI](https://langfuse.com/docs/api-and-data-platform/features/cli)** (`npx langfuse-cli api `) to read or write traces, prompts, datasets, and scores from the terminal. Found an error in these docs? Please open an issue at .